1

Kenya’s Experience with Power-Sharing

Jeremy Horowitz Assistant Professor Dartmouth College

This Draft: October 11, 2013

2

Introduction

This chapter examines the power-sharing agreement that was implemented in

Kenya from 2008 to 2013. It was adopted in responses to widespread ethnic violence

triggered by a disputed presidential election in late 2007 and expired with the country’s

subsequent multiparty election in 2013. The power-sharing accord had the immediate

effect of ending the spiral of communal violence that threatened to lead the country into a

full-blown civil war. The institutions, however, did little to encourage elites to implement

substantive reforms aimed at ameliorating the longer-term sources of conflict. To explain

these outcomes, the chapter argues that in Kenya power-sharing contributed to the

cessation of violence through two related mechanisms. First, by coopting opposition

leaders, the power-sharing accord removed incentives for opposition politicians to

encourage or support the post-election violence in order to enhance their bargaining

position with the incumbent government. Second, the incorporation of opposition figures

provided an important signal to their supporters, mollifying the fears of exclusion and the

inter-communal resentments that contributed to the eruption of violence. Yet, the power-

sharing institutions did little to incentivize deeper reforms among political leaders with

vested interests that ran counter to change. In sum, the Kenyan case suggests that power-

sharing can serve as an effective strategy for addressing outbreaks of ethnic violence, but

is less likely to generate incentives for more thorough-going reform in settings where

political elites prefer the status quo.

The structure of this chapter is as follows. The first section describes the sources

of instability in Kenya, focusing on the political salience of ethnic divisions. The second

section describes the 2008 power-sharing agreement. The third section looks at its effects,

3

evaluating its impact against the primary goals articulated during the negotiations

process, namely stopping the violence and addressing deeper sources of ethnic tension.

The fourth section examines the potential for renewed conflict.

1. Challenges to Political Stability in Kenya

This section provides a brief overview of the threats to stability in Kenya. It

distinguishes between long-term sources of tension that stem from the politicization of

ethnicity since independence and the short-term factors that contributed to the outbreak of

violence after the 2007 election.

The Politicization of Ethnic Differences

The divisiveness of ethnic cleavages in Kenyan is in large part the result of a

history of ethnic favoritism by successive Kenyan leaders.1 It is widely believed that the

first President, Jomo Kenyatta, favored his own ethnic community, the Kikuyu, in a

variety of ways. After independence, Kenyatta appointed many trusted members of his

own Kikuyu ethnic group, particularly those from his home area of Kiambuu in the

Central Province, to high-level positions in the Cabinet and other state agencies (Throup

1987). Moreover, the land redistribution scheme, undertaken shortly after independence,

was seen as disproportionately benefiting Kikuyus (Kyle 1999). Because Kikuyus were

1 While this section draws primarily from anecdotal and historical accounts, there is a growing body of empirical research that finds evidence of ethnic favoritism in public service delivery (e.g., Franck and Rainer 2012; Kramon and Posner 2012; Burgess et al., 2013; Jablonski forthcoming). Moreover, public opinion surveys routinely find that Kenyans believe ethnic favoritism to be endemic to Kenyan politics (e.g., Ross 1975). In my own survey, conducted in November-December 2012 (N=1,246), an overwhelming majority of respondents indicated that favoritism is common. When asked how much of the time government leaders favor their own ethnic groups, 52% said “almost always” and 31% said “some of the time.” Only 12% said “rarely” and 3% “never.”

4

relatively better off than other ethnic groups at independence, they were able to take

advantage of the sale of highly productive land that was purchased from departing white

settlers by the government and then resold on a “willing-buyer, willing-seller” basis. In

addition, it is widely believed that Kenyatta handed out prime tracts of land and many

business concessions to co-ethnic supporters (Kanyinga 2009; Kamugi 2009; Harbeson

2012).

After Kenyatta’s death in 1978, the Vice President, Daniel arap Moi, a Kalenjin

from the Rift Valley, assumed to the presidency. Like Kenyatta, Moi relied on an inner

circle drawn largely from his own ethnic group, the Kalenjin (Kanyinga 2007, Throup

1987). Moreover, Moi promoted co-ethnics within the government administration,

parastatals, the military, and the police, perpetuating the sense that state largesse was

distributed along ethnic lines. The advance of Kalenjins within government structures

came in many cases at the expense of Kikuyus (Kanyinga 2007).

In 2002, Mwai Kibaki, a Kikuyu from Central Province, became Kenya’s third

President. Kibaki came to power at the head of a diverse coalition party, the National

Alliance Rainbow Coalition (Ndegwa 2003). While Kibaki was elected on a promise to

end tribalism, in practice, Kibaki, like those before him, was seen as favoring his own

group, particularly in promotions to top Cabinet and administration positions (Africa

Confidential 2003, Barkan 2008).

A second factor behind the politicization of ethnic differences in Kenya is the

party system, which has historically reflected and reinforced ethnic cleavages. Prior to

independence, competing parties emerged as champions for different ethnic blocs,

turning early electoral contests into an “us-against-them” battle for control of the state

5

and its resources. The first party to form in the early 1960s was the Kenya African

National Union (KANU), which initially brought together leaders from across the ethnic

spectrum, united in the goal of gaining independence from Britain. However, within a

short time, leaders from several smaller ethnic groups split from KANU after becoming

convinced that they would never play a major role within the party. The break-away

faction, headed by leaders from the Kalenjin, Maasai, and coastal groups, formed the

Kenya African Democratic Union (KADU). Central to KADU’s platform was a demand

for a federal constitution that would give ethnic groups greater control over their own

affairs and limit perceived domination by larger tribes. This position resonated with

Kenyans in the Rift Valley and on the coast who were worried about Kikuyu and Luo

“migrants” snatching up prime land in these areas (Sanger and Nottingham 1964). In the

two national elections before independence – held in 1960 and 1963 – ethnic identities

played an important role in determining voting behavior, with many voters lining up

behind their “ethnic champions” (Bennett and Rosberg 1961, Sanger and Nottingham

1964).

Shortly after independence in 1963, Kenya became a de-facto single party state

under the KANU government, which had emerged victorious in the 1963 election. For

the subsequent two decades, opposition parties did not exist. When multi-party

competition was again legalized in 1991, parties once again formed around ethnic blocs

(Throup and Hornsby 1998). In the 1992 and 1997 elections, the opposition parties

garnered support from Kikuyu and Luo areas, while the incumbent party, headed by Moi,

maintained its strongest support among Moi’s own Kalenjin group and among

pastoralists from the Rift Valley (Barkan 1993, Barkan and Ng’ethe 1998). In the 2002

6

election, opposition parties united behind Kibaki, drawing support from most ethnic

communities, other than the Kalenjin who by and large remained loyal to KANU

(Ndegwa 2003).

A third factor, which is both a reflection of the politicization of ethnic differences

and which has sharpened such divisions, is a history of episodic violence associated with

elections. In the multi-party era since 1991, significant ethnic violence has taken place in

three election rounds (1992, 1997, and 2007). Much of the responsibility for this violence

is thought to rest with politicians who have exploited latent communal tensions for

electoral gains, fueling divisions between groups. For example, prior to the 1992 election,

Kalenjin politicians in the Rift Valley used fear tactics to mobilize support within their

communities, claiming that Kikuyus would seek retribution against Kalenjins if they

came to power (HRW 1993, Throup and Hornsby 1998). Leaders called on Kalenjins to

defend their ethnic homelands and to chase Kikuyus out of the area. Studies of this period

have found that top KANU politicians funded groups of “Kalenjin warriors” that attacked

and killed Kikuyus living in the area. Reports estimate that 1,500 people died in these

clashes and another 300,000 were chased away from their homes, mostly in the Rift

Valley, during the months before and after the 1992 election (HRW 1993). A similar

dynamic was seen in the lead-up to the 1997 election, with violence this time occurring

both in the Rift Valley and also in some ethnically-mixed areas along Kenya’s coast

(KHRC 1997; KHRC 1998). The violence that erupted following the disputed election in

2007, described below, should therefore be seen as a part of a general pattern, not an

isolated event.

7

The 2007 Election

The 2007 presidential election featured two main candidates, drawn from different

ethnic communities. The incumbent President Mwai Kibaki, a Kikuyu from Central

Province, headed the Party of National Unity (PNU). His main challenge came from

Raila Odinga, a Luo from Nyanza Province in southwestern Kenyan, who led the Orange

Democratic Movement (ODM). The election also included Kalonzo Musyoka, a Kamba

from Eastern Province, who headed the Orange Democratic Movement-Kenya (ODM-K).

By the time the 2007 election campaigns got under way, Kenyans were already

severely polarized. Despite real achievements made by the Kibaki government in terms

of infrastructure, economic growth, education, and expanded political freedoms, many

Kenyans felt Kibaki had failed to live up the promises he offered in 2002. As mentioned,

Kibaki had come to power on a promise to end tribalism. Yet, after taking office he

reneged on a pre-election commitment to create a Prime Minister position that was to be

filled by Odinga. To many Luos, this was seen as a betrayal. Moreover, during Kibaki’s

first two years, a divisive constitutional review process took place, culminating in an

acrimonious referendum that pit ethnic groups against each other (Lynch 2006, KNCHR

2006). Kibaki’s administration also became embroiled in a number of conspicuous

corruption scandals, betraying his promise to end the pervasive corruption of the Moi era.

But the Achilles heel for Kibaki was the perception that he had favored his own

Kikuyu ethnic group at the expense of others. A survey conducted in mid-October 2007,

about two months before the election, found that the majority (57%) of Kenyans thought

that Kibaki’s government served the interests of certain ethnic groups at the expense of

8

others.2 Among ethnic groups aligned with the opposition, large majorities believed that

Kibaki favored his own community. For example, 90% of Luos and 79% of Kalenjins

thought that the Kibaki government favored some groups over others. And among those

who felt that the government favored certain groups over others, the vast majority (91%)

said that the government favored Kikuyus.

In the election campaigns, the opposition parties – ODM and ODM-Kenya –

exploited these perceptions, using emotional appeals that tapped into a sense of injustice

and resentment (Horowitz 2012; Barkan 2008). As evidence, opposition candidates

claimed that Kibaki had disproportionately appointed Kikuyus to top positions in the

Cabinet and the government administration. Moreover, the opposition charged that

Kibaki unfairly favored the Kikuyu region – Central Province – in the distribution of

funds for social services and infrastructure, while neglecting other parts of the country.

At the same time, the opposition alleged that Kibaki had callously persecuted certain

groups, for example by evicting Kalenjin squatters from the Mau Forest, a protected

watershed, and deporting Muslims from coastal regions. Opposition appeals played on

long-standing grievances against Kikuyus, who have historically been better off

economically than other groups in Kenya. As expected, voting closely followed ethnic

lines, with voters from several groups lining up en masse on opposite sides of the ethno-

political divide (Gibson and Long 2009).

A Disputed Election as Trigger for Post-Election Violence

2 Data comes from a survey conducted by the Steadman Group (n=2,718) in mid-October, 2007.

9

The national election took place on December 27, 2007. Over the next few days,

the Electoral Commission of Kenya (ECK) failed to address concerns about possible

manipulation of the vote counting process (Throup 2008). The official figures showed

Kibaki ahead of Odinga by a sizable margin, but ODM leaders argued that the vote

counts from several areas had been inflated to favor Kibaki after the polls closed. Despite

lingering concerns about the integrity of the tally, the ECK Chairman, Samuel Kivuitu,

announced Kibaki as the official winner in the early afternoon of December 30. Less than

two hours later, Kibaki was sworn in at a rapidly assembled inauguration ceremony.

Minutes later, the violence began.

Over the next two months at least 1,000 people were killed and another 300,000

were chased from their homes (ICG 2008; KNCHR 2008; CIPEV 2008; Anderson and

Lochery 2008). While much of the violence initially appeared to be spontaneous rioting,

it quickly became clear that Kibaki’s ethnic group, the Kikuyu, were being targeted.

Kikuyus living in ethnically-mixed areas had their houses burned down and their

businesses destroyed. Many fled their homes and sought protection in police stations and

other government buildings. The brutality of the clashes was conveyed to the outside

world on January 1, 2008 when an angry mob burned down a church in Eldoret in which

dozens of people had sought refuge. Kikuyus fearing for their lives headed toward the

Central Province, the region where the ethnic group is concentrated. At roadblocks set up

along the main highways, gangs pulled people from their cars, beating and killing those

with the “wrong” identity. Shortly thereafter, reprisal killings began, as Kikuyu militias

sought to stem the violence and exact revenge. As the violence escalated, observers

worried that the clashes would spiral into a much larger conflict.

10

2. The 2008 Power-Sharing Agreement

As violence between ethnic communities escalated in the weeks after the election,

a long list of foreign dignitaries attempted to restore stability and resolve the political

impasse between PNU and ODM. This list included Bishop Desmond Tutu; John Kufuor,

President of Ghana and out-going head of the African Union (AU); UN Secretary

General Ban Ki-Moon; and US Secretary of State Condoleezza Rice; among others.

Foreign heads of state and international donors put pressure on the two main political

parties to bring an end to the violence (Lindenmayer and Kaye 2009; Brown 2009;

Kanyinga and Walker 2013). Little progress was made, however, prior to the arrival of

the negotiating team headed by former UN Secretary General Kofi Annan on January 23,

2008.

The power-sharing negotiations began with a wide gulf in the positions of the two

sides (for detailed accounts of the negotiations see Lindenmayer and Kaye 2009 and

Griffiths n.d.). The opposition party, ODM, believed that it had rightly won the

presidential election. ODM’s position was that an interim government should be created

for a period of six months, after which time the presidential election would be re-run. On

the other side, PNU was unwilling to consider a re-run of the election, and argued instead

that ODM should submit any complaints regarding election fraud to the courts. Once it

became clear that a re-run of the election was unlikely, the negotiations shifted to seeking

terms for a power-sharing arrangement that would allow the two parties to govern jointly.

As the negotiations proceeded, ODM advanced several core demands. First, it

sought the creation of a Prime Minister (PM), who would serve as Head of Government

11

and have the authority to appoint and remove government officers, including Cabinet

members. Second, ODM demanded that Cabinet portfolios be divided equally and that

high-profile ministries – e.g., Interior and Finance – be shared between the two parties.

Third, ODM called for proportionality at all levels of government. Lastly, ODM sought

to have the deal entrenched in the constitution. On the other side, PNU demanded that the

President remain as Head of Government and retain the authority to determine the

composition of the Cabinet. PNU’s negotiators pushed for a Prime Minister that would

oversee the ministries but would not have executive functions related to hiring or firing.

At several points during the negotiations, which unfolded over five weeks, the

intransigence of the two sides threatened to derail the bargaining. Annan’s adroit

management was of considerable importance in maintaining progress. Early on Annan

ended the acrimonious war of words that was being waged in the media by the two

parties. At a critical point, Annan sequestered the negotiators for several days in Tsavo,

one of Kenya’s wildlife preserves, to remove the process from the limelight in Nairobi,

the nation’s capital. And in the last stages when the negotiations were stalled over key

details, Annan bypassed the negotiation teams appointed by each party and appealed

directly to the principals, the heads of the two parties. This move marginalized hardliners

within the two camps, resulting in the final agreement.

The final deal, which was signed on February 28, was limited in scope.3 As

shown in Table 1, the accord called for the establishment of a coalition government that

would include PNU and ODM as equal partners. The position of Prime Minister would

be created and would be filled by the party or coalition with the largest number of seats in

3 “Agreement on the Principles of Partnership of the Coalition Government.” February 28, 2008. The agreement was entrenched in the constitution through adoption by Parliament of the National Accord and Reconciliation Bill, 2008.

12

Parliament. In practice, this meant that ODM, which held the largest share of seats, would

be able to nominate Odinga to fill the position. Two Deputy Prime Ministers would be

created, one to be filled by ODM and the other by PNU. In appointing Cabinet ministers,

the principle of “portfolio balance” would be followed, whereby the number of seats

would be assigned in proportion to the parties’ strength in parliament.

[TABLE 1 ABOUT HERE]

To safeguard against the possibility of reneging, guarantees were provided to

ODM, which as the opposition party was in the weaker position. It would only be

possible to remove the Prime Minister through a vote of no confidence passed by a

majority in parliament. Given that ODM held the largest share of Parliamentary seats,

this provision meant that it be difficult to remove the PM without ODM’s consent.

Similarly, Cabinet Ministers would be removed from office only with concurrence from

both parties. And the agreement would be entrenched in the constitution through a bill

passed by Parliament.

Other aspects of the agreement remained vague. Regarding the powers of the

Prime Minister, the accord stipulated that the PM would “co-ordinate and supervise”

affairs of government, without specifying exactly what power and authority would be

vested in the position and how executive powers would be divided between the PM and

the President. While it was clear that that the agreement was temporary in nature, the

accord left unresolved the question of how long the coalition government would last and

what would happen if it collapsed. The agreement stipulated that the coalition could be

13

dissolved if the parliament were dissolved, if the parties agreed in writing, or if one

coalition partner withdrew from the coalition. However, the agreement did not specify

what would happen if the coalition were to collapse before the next election scheduled for

2012. In the event, the coalition survived through the full term, expiring at the subsequent

election, which after being delayed was held in March 2013.

In addition to the creation of a power-sharing coalition, the parties agreed to address

the root sources of conflict, the so-called “Agenda Four” items listed on the initial agenda

for the negotiations process. These included constitutional reform, regional economic

imbalances, land reform, and poverty, among others.4 In this spirit, the parties

subsequently agreed to establish a set of investigatory committees charged with

examining fraud during the election, the causes of the post-election violence, and the

sources of long-standing grievances in Kenyan society.

The first of the three committees to get underway, the Independent Review

Committee, was charged with looking at electoral fraud during the election. The

committee, headed by former South African judge Johann Kriegler, issued the so-called

“Kriegler Report” in September 2008. The report ruled that it was not possible to conduct

a re-count of the ballots in order to determine the actual winner of the 2007 election. The

main recommendations of the report related to the overhaul of the Electoral Commission

of Kenya, and particularly the adoption of measures to strengthen the body’s

independence.

The second investigatory committee, the Commission of Inquiry on Post-Election

Violence, was mandated to examine the sources of the post-election violence. The

4 “Agenda for the Kenyan National Dialog and Reconciliation.” Accessed online at http://www.irinnews.org/pdf/Annotated%20Agenda%20Ver%202.pdf on October 22, 2008.

14

committee, which was headed by High Court Justice Philip Waki, submitted its findings

to Annan on October 17, 2008. The report provided a detailed account of the post-

election violence and included a list of alleged perpetrators of the violence, implicating

several sitting Ministers and Members of Parliament (MPs). The Commission called on

the Coalition government to establish a tribunal to bring the alleged perpetrators to trial.

However, within days of the report’s release, it became clear that there was weak support

for its implementation among many top leaders in the government. President Kibaki

urged Kenyan to be cautious in pushing for prosecutions, arguing that doing so might

undermine peace and stability (Daily Nation, 10/21/2008). ODM was initially divided

over the report. While Prime Minister Odinga and some top leaders called for its full

implementation, the party announced its opposition to the report after several senior MPs

from the Rift Valley – William Ruto, Henry Kosgei, and William ole Ntimama – opposed

it (Daily Nation, 10/24/2008). With political leaders failing to launch a domestic tribunal,

Justice Waki, made good on his threat to deliver the list of alleged perpetrators to the

International Criminal Court, which subsequently opened investigations in Kenya in

2010.

Last, parliament created the Truth, Justice, and Reconciliation Commission

(TJRC) in 2008. The body was charged with a broad mandate to investigate all forms of

historical injustice committed between the nation’s independence in 1963 and 2008 and

to recommend appropriate actions or reforms. Yet, the eventual composition of the body

insured that it would have little moral authority. As its chair, President Kibaki appointed

Bethuel Kiplagat, a former high-ranking member of the government and diplomatic

corps, with a controversial record of alleged improprieties related to land theft and

15

political violence. Whether by intention or not, Kiplagat’s appointment hamstrung the

commission, leading to the active revolt by other commission members and judicial

proceedings seeking his removal. The controversy over his involvement on the

commission served both to distract attention from the Commission’s work and to

undermine its legitimacy (Hansen 2013). The final report, issued in May 2013, contained

little new information.

3. Effects of the Power-Sharing Agreement

Did Kenya’s power-sharing agreement work? In answering this question, I

evaluate the agreement against the two main goals set forth in the negotiations that led to

its creation, namely ending the post-election violence and addressing the underlying

conditions that contributed to it. I focus primarily on the key institutional innovation

contained in the agreement, the coalition government that brought together elites from the

two main political factions, and argue that the power-sharing arrangements had strong

positive effects on the violence but less substantial effects on the long-term sources of

tension.

Ending the Violence

By any measure the power-sharing agreement succeeded in achieving the primary

goal of ending the conflict and restoring stability. How did it do so? Answering this

question is important both for making sense of Kenya’s unique experience and for

contributing to larger debates about when and how power-sharing can be an effective

remedy to ethnic violence.

16

Debates about power-sharing as a solution to ethnic violence typically focus on

civil war settings in which one or more armed group challenges the state militarily. In this

context, one way in which power-sharing agreements are thought to help restore peace is

by making negotiated settlements credible. As Walter (1997) has argued, the main

impediment to negotiated settlements is often the concern by rebel forces that they will be

persecuted by state powers after laying down their arms, regardless of the guarantees

provided in the peace accord. Building on this logic, Gates et al. (2013) suggest that

power-sharing agreements contribute to peace only when they include institutional

protections for rebel groups and the communities that support them.

While this theoretical framework may be useful for understanding the effects of

power-sharing in civil wars, it is less useful in cases like Kenya where the violence was

more informal in nature. Anecdotal accounts suggest that the perpetrators were loosely

organized groups, sometimes using violence against other members of their localities

spontaneously and sometimes with coordination by local community leaders,

businessmen, and politicians (IGC 2008; Njogu 2009; KNCHR 2008; CIPEV 2008).

While reports suggest that more formally organized groups, in particular the Mungiki, did

assume a more central role as the violence escalated, the conflict was not one between

organized fighting forces and the state. This is important for two reasons. First, it meant

that at the elite level, the main parties to the negotiations were not military leaders who

might be preoccupied with concerns about their security in the post-conflict period.

Rather, they were politicians, who in all likelihood were concerned first and foremost

with access to positions of power within the next government. Second, the informal

nature of the violence meant that post-conflict security concerns would have also been

17

less relevant for the foot soldiers, many of whom would have likely been able to avoid

persecution for their crimes with relative ease. Thus, the “critical barrier” to peace in

Kenya was not security concerns; rather the key challenge was to find a way to break the

spiral of violence before it escalated further.

In this context, the power-sharing accord worked through two related

mechanisms. First, it served to co-opt opposition leaders in ODM, and in so doing

reduced incentives for these leaders to encourage further violence. As an IGC report

claimed, senior ODM leaders saw the violence as a useful bargaining tool during the

negotiations. The report argued that “ODM believes that if international mediation fails,

its only protection against repression and hope for a settlement will be its capacity to

raise the stakes through violence” (ICG 2008, p. 24). While it is unclear whether ODM

leaders exerted much direct control over the violence, the successful end to the

negotiation process eliminated any incentives to support the violence to enhance their

bargaining position with PNU. It is noteworthy that while Odinga called for peaceful

protests during the negotiations, his most forceful appeal for peace came only at the

conclusion of the power-sharing accord, when he asked his supporters to move from the

“phase of confrontation to the beginning of cooperation” (BBC, 2/28/2008).

Second, the power-sharing agreement served an important signaling function for

opposition supporters. In previous sections of this chapter I argued that the violence was

fueled in part by perceptions that one community – the Kikuyu – had unfairly benefited

from power while others had been excluded. The alleged theft of the 2007 election by

PNU – a party led by Kikuyu politicians – brought to a head resentment against the larger

community. In coopting opposition leaders, the power-sharing agreement signaled that

18

communities aligned with the opposition would not be excluded from the benefits of

power after the election. In Kenya, it is widely believed that gaining a fair share of state-

controlled resources for one’s ethnic group depends on electing leaders from one’s

community to positions of power. By creating the position of Prime Minister, which was

to be filled by Odinga, and by institutionalizing the principle of proportionality in the

cabinet and the bureaucracy, Kenya’s power-sharing agreement signaled that state funds

would be distributed equitably across groups. By doing so, it addressed a central

grievance that motivated anti-Kikuyu violence in the wake of the 2007 election.

It is important to note that the agreement offered few provisions to allay possible

concerns about the credibility of the deal. In particular, there were no guarantees that the

agreement would insure equitable treatment across ethnic communities in the event that

PNU might seek to use its control of the presidency to favor its own constituents. Nor did

the agreement put in place the kinds of institutional protections that Gates et al. (2013)

argue are critical for power-sharing to be effective over time. These observations

strengthen the point that Kenya’s power-sharing agreement worked mainly by coopting

opposition elites and by helping to mitigate the anger and frustration that fueled the

violence, consistent with conventional arguments made by power-sharing advocates (e.g.,

Lijphart 1997; Sisk 1996). While Gates et al. 2013 fail to find any systematic relation

between inclusive power-sharing arrangements and stability in a global sample of

countries, the Kenyan case suggests that in some settings inclusion may be a useful

institutional remedy.

Addressing Long-Term Sources of Conflict

19

While Kenya’s power-sharing agreement had an immediate positive effect on the

post-election violence, the agreement did little to generate incentives for elites to address

the deeper sources of instability. The bright spot on the reform agenda has been the

ratification of a new constitution in 2010. However, it is debatable whether this success

can be attributed to the power-sharing institutions. As argued below, key political leaders

had individual incentives that may have inclined them to cooperate in producing the new

constitution, and conflict, not cooperation, has been the more typical dynamic between

the coalition partners. Moreover, little or no progress has been made in other critical

areas, such as land reform, bringing to trial those responsible for the election violence, or

altering the zero-sum nature of electoral competition. I review these aspects in turn,

arguing that while the power-sharing agreement may have contributed to the passage of

the new constitution, the agreement did not diminish elite resistance to more thorough-

going reforms.

The 2010 constitution is perhaps the crowning achievement of the coalition

government. Marking the culmination of nearly two decades of efforts to reform key

aspects of the Kenyan political system, the constitution has been widely heralded as a

major milestone in Kenya’s journey toward democracy (Barkan and Mutua 2010;

Kramon and Posner 2011). Notable in this regard are provisions that devolve power and

resources to county-level government, increase legislative oversight of the president, and

enhance individual rights.

To what extent did the power-sharing agreement contribute to the new

constitution? Two instances of elite cooperation were critical to its passage, and the

power-sharing agreement might plausibly have increased incentives for key leaders to

20

cooperate. As in the previous attempt to draft a new constitution in 2004, top party

leaders held widely divergent preference on critical provision (Kramon and Posner 2011;

Amandi 2009; Murray n.d.). A key priority for Odinga and ODM was devolution, which

was strongly opposed by PNU leaders. Likewise, Odinga preferred a parliamentary

system with a Prime Minister to be chosen by parliament, while the PNU contingent

favored the retention of a presidential system with a directly-elected chief executive. In

2004, political leaders failed to reach compromise on similar issues, leading to the break-

down of the reform effort and the ultimate rejection of a government-sponsored draft in a

national referendum. By contrast, in 2010 a compromise was reached, whereby ODM

abandoned its demands for a shared executive and PNU dropped its opposition to

devolution, paving the way for the final version of the draft, which was approved by

voters in a subsequent referendum. Second, in sharp contrast to the 2005 referendum, in

which Odinga led the charge against the proposed constitution, Odinga and Kibaki joined

forces to persuade the Kenyan public to support the 2010 referendum. The alliance

between Odinga and Kibaki is all the more notable in that key members within their

respective parties were either lukewarm on the draft or in outright opposition to it. Given

the influence top leaders exercise on public opinion, it is doubtful that the 2010

referendum would have passed had Odinga and Kibaki not come together to support it.

It is not implausible to think that the power-sharing agreement encouraged

cooperation and compromise in drafting and ratifying the constitution. In its absence,

Odinga and ODM might have had stronger incentives to oppose any cooperation with

Kibaki and PNU, both as a matter of principal and in order to avoid sharing credit for a

popular initiative. However, as part of a coalition government, Odinga and ODM bore a

21

greater burden of blame for inaction and stood to reap greater benefits from satisfying

public demands for reform.

Yet, observers have more typically viewed the coalition government as an

impediment to reform (e.g., Amandi 2009; Cheesemen and Tendi 2010). And it is

important to note that both Odinga and Kibaki may have had personal motivations for

supporting constitutional reform. Kanyinga and Long (2012) note that there was

considerable support for reform among the Kenyan public. They speculate that for

Odinga it may have been important to deliver on his campaign promise to revise the

constitution, and that doing so would position him well for another run at the presidency.

Likewise, the authors hypothesize that although Kibaki was barred from seeking a third

term in office, concerns about his legacy, which had been greatly tarnished by the 2007

election, may have motivated him to support the reform initiative. Moreover, the key

compromises on the content of the reform may have also stemmed from private interests.

For Odinga, the 2007 election demonstrated that he was a viable contender for the

presidency, and his desire to capture this post in 2012 may well have led him to abandon

his insistence on a parliamentary form of government (Kramon and Posner 2011). Thus,

while the power-sharing agreement may have reduced the opposition’s incentives to

oppose collaboration with PNU, it is difficult to separate these institutional effects from

other incentives that may have inclined also key actors to support the constitutional

reform effort.

It is noteworthy, moreover, that these episodes of cooperation stand in sharp

contrast to the more acrimonious and divisive relations that typified the coalition

government throughout its lifespan. From its inception the coalition government was

22

hampered by disputes over basic procedural issues within the legislature. The first major

task under the accord was the assignment of MPs to Cabinet posts in the new coalition

government. While both parties were prepared to honor the principal of portfolio balance,

initial negotiations deadlocked over which positions would go to each party. Both sides

sought control of the key ministries, such as Finance, Internal Security, and Public

Service, among others. The stalemate, which dragged on for several weeks, threatened to

undermine the accord. In frustration over the delays and PNU’s perceived hardline

position, ODM supporters in Kisumu and Nairobi returned to the streets for riots in April

2008 (Daily Nation, 4/2/2008). Equally frustrated with the impasse, the European Union

threatened to cut off aid if a deal was not reached (Daily Nation, 4/9/2008). After several

tense weeks of negotiations, the parties finally called on Annan to help broker a deal. The

final agreement, which was announced on April 13, allowed PNU to keep Internal

Security and Finance, while ODM gained other key ministries, including Public Service,

Local Government, and Agriculture (Daily Nation, 4/14/2008). In May 2009 the two

parties again deadlocked over the question of which party would fill the position of

Leader of Government Business within the House, a stalemate that was only broken when

the Speaker of the House stepped in and assumed the position himself (Daily Nation,

6/4/2009). The parties also came to an impasse on the assignment of members to the

twenty-seven committees within the House. As a result of these procedural tussles, all

legislative business was halted for several months in 2009 (Standard, 6/14/2009).

The coalition government was also characterized by persistent accusations of foul

play with regard to the core provisions of the agreement. In March 2009, for example,

ODM’s leadership called for a renegotiation of the deal, alleging that the party had been

23

shut out of a series of important decisions that had been made unilaterally by PNU. ODM

cited in particular the unilateral appointment of staff to several government agencies.

ODM’s Secretary General, Peter Anyang Nyong’o, summarized the party’s position as

follows: “The Coalition Government has fallen far short of implementing the National

Accord in letter and spirit, thereby shortchanging our Party in appointments to the civil

service, public corporations, and the security services" (Daily Nation, 3/3/2009). Taken

together, these instances suggest that the power-sharing accord did little to enhance

incentives for elite cooperation overall.

Likewise, the power-sharing institutions did little to diminish elite resistance to

addressing the more fundamental causes of conflict in Kenya. Aside from the passage of

a new constitution, little progress was made on any of the Agenda Four items. It is not

difficult to see why: members of the political elite had private interests that ran counter to

reform. In this context, the creation of a grand coalition drawn from the existing elite did

not lead to substantive reform because it did not create new rewards for reform or alter

the composition of the policy-making elite. As a result, continuity, rather than change,

has been the outcome in most areas (Cheeseman and Tendi 2010; Hansen 2013).

The failure to address deeper sources of conflict can be seen first in the decision

not to pursue domestic legal proceedings against those responsible for the post-election

violence. As many have noted, violence has become a routine feature of elections in

Kenya in part because the judicial system has so far failed to take action against those

who commit atrocities (KNCHR 2008; Kamugi 2009). Thus, ending the impunity

enjoyed by the brokers of violence has been identified as a core priority for reducing

future bloodshed (KNCHR 2008; CIPEV 2008). In the aftermath of the 2007 violence,

24

however, Kenya’s political leaders showed little interest in punishing either the

organizers or the foot-soldiers. As noted, prominent leaders, including President Kibaki,

argued that a judicial approach would further divide the country, jeopardizing the fragile

peace that had been achieved. Instead, the President called for a blanket amnesty. The

decision by the International Criminal Court to investigate the violence and ultimately to

open a case against several top leaders briefly provided an impetus for a debate about

whether to launch a domestic tribunal. Once again, however, the political elite failed to

take action. Given that many high-ranking political leaders are thought to have been

involved in orchestrating the violence, key actors in the coalition government had strong

incentives to resist a judicial response (Brown and Sriram 2012; Hansen 2013). By

creating an oversized coalition drawn from the existing political elite, the power-sharing

agreement ensured that these actors would be well positioned to block the creation of a

local tribunal. While it is perhaps unlikely that such a tribunal would have materialized in

the absence of the power-sharing accord, it is clear that the institutions of power-sharing

did little to create incentives for reform in this area.

Political leaders have also shown little enthusiasm for efforts to investigate or

address historical injustices related to inequities in regional development or access to

land. As noted above, the TJRC was hampered by disputes about its leadership and

ultimately failed to unearth substantial new details about past injustices. More recently,

the National Land Commission – a product of the 2010 constitution – was formed with a

similar mandate to examine historical improprieties particularly with regard to land and

to offer recommendations for reform. Yet, like the TJRC, the Land Commission is

unlikely to have much impact, as it is mandated only to offer recommendations and lacks

25

the authority to initiate reforms on its own (Business Daily, 3/24/2013). The lack of

enthusiasm for more substantial action on historical injustices is not surprising given that

many senior political leader are thought to have been involved in or benefited from past

abuses of power (KNCHR 2008; CIPEV 2008). While the history of inaction and

obfuscation, particularly on land, suggests that little progress would have occurred in the

absence of the power-sharing accord, it is again clear that the creation of a grand coalition

did little to incentivize reform.

Finally, progress in altering the polarizing dynamics of electoral competition has

been modest. The 2010 constitution contained a number of provisions designed to

encourage broad-based parties that appeal beyond their core ethno-regional

constituencies. Notably, the presidential winner is now required to garner at least 50% of

the national vote (previously a plurality was sufficient) and to win at least half the votes

in 25% of Kenya’s 47 counties (previously the winner was required to gain at least 25%

of the vote in five of Kenya’s eight Provinces). Yet, if the 2013 election is any guide,

such provisions will not fundamentally alter the basic zero-sum nature of politics. In the

2013 election most Kikuyus and Kalenjins supported the coalition headed by Uhuru

Kenyatta and William Ruto while most Luos and Kambas lined up behind the coalition

headed by Odinga and his running mate Kalonzo Musyoka.5 These voter alignments gave

the race an “us-against-them” quality in which opposing ethnic groupings once again

squared off in competition for power. Given that politicians benefit from the tendency by

voters to engage in ethnic bloc voting, it is unlikely that additional institutional reforms –

5 Survey data from a poll conducted by the author shortly before the 2013 election showed that 94% of Kikuyus and 90% of Kalenjins preferred Kenyatta to Odinga while 96% of Luos and 85% of Kambas preferred Odinga. Data come from a nationally-representative poll (excluding Northeastern Province) with a sample of 829 respondents conducted between February 12 and March 3, 2013.

26

such as the alternative vote or other preference-based systems – will be adopted.

Moreover, little action had been taken by political leaders to reduce the use of hate

speech during campaigns. Despite the creation of the National Cohesion and Integration

Commission, which is charged with monitoring and bringing cases against those who

engage in hate speech, and the inclusion of Article 49, which makes hate speech and

incitement illegal, in the new constitution, monitoring and enforcement of campaign

improprieties remains weak (Institute for War and Peace Reporting, 7/11/2013). Again,

given that members of the political elite may well have run afoul of such laws in the past,

these actors are unlikely to push for more comprehensive reform. And while reform

would have been unlikely in the absence of power-sharing institutions, this instance again

shows that the power-sharing institutions did not mitigate elite resistance.

In sum, while the power-sharing agreement in Kenya played a critical role in

interrupting the spiral of post-election violence, it did little to encourage political leaders

to address deeper sources of conflict. These conclusions are reflected in public opinion

data. In a national survey conducted in February 2013, just before the agreement’s

expiration, 73% of respondents approved of the power-sharing agreement, with 86% of

respondents agreeing that “the power-sharing agreement helped Kenya avoid additional

violence.”6 Smaller shares of respondents, however, agreed that the agreement helped to

rebuild trust between ethnic communities (56%) or to resolve long-standing tensions

between ethnic groups (61%). And only 39% agreed that power-sharing should be used in

the future when there are disputes about the outcome of a presidential election,

6 Data come a public opinion survey (N=829) carried out February 12 and March 8, 2013. The survey was nationally representative but excluded Northeastern Province because of security concerns.

27

suggesting that while many Kenyans view the agreement as an appropriate stop-gap

measure, most prefer that power-sharing not become routine.

4. Potential for Future Conflict

While the power-sharing agreement restored stability in Kenya, potential for

future conflict remains. First and foremost, long-standing grievances over land

distribution, inequalities in wealth, and perceived historical injustices have not been

addressed. The election of Uhuru Kenyatta – son of Kenya’s first President Jomo

Kenyatta – as President and William Ruto as Deputy President in the 2013 election

suggest that there will be little support in coming years for examining improprieties in

government land dealing or undertaking substantial reform to address grievances over

access. The Kenyatta family is thought to have been among the largest beneficiary of

government land distribution schemes after independence, possibly owning as much half

a million acres (East African Standard, 10/1/2004). Likewise, while Ruto is of more

modest parentage, he is thought to have benefited from land grabbing in the Rift Valley

and was recently ordered by the High Court to hand over a 100-acre farm that he is

thought to have obtained illegally during the 2008 post-election violence (IOL News,

6/28/2013).

Equally troubling is the culture of impunity that has long existed in Kenya,

whereby perpetrators of violence are not held to account for their actions. The periodic

cycle of violence during elections now spans several election rounds, dating back to

1991. Many of the same politicians have been repeatedly implicated in the violence. For

these actors, violence and polarization are useful electoral strategies. The failure to

28

punish political leaders who encourage violence means that they will have little incentive

to refrain from such behavior in the future.

Finally, elections continue to serve as flashpoints for inter-communal tensions. As

noted, constitutional provisions designed to alter the zero-sum nature of electoral

competition have done little to change the basic dynamics of party politics. Tensions in

the Rift Valley, a major hotspot in the 2008 post-election violence, were muted in the

2013 election as a result of the political alliance between Kenyatta (a Kikuyu) and Ruto

(a Kalenjin). And Kenyatta’s victory may help to allay fears of marginalization among

Kalenjin voters in the short term. Yet, the alliance does nothing to diminish similar fears

among other groups, like the Luo and Kamba, that supported Odinga in large numbers in

2013.

Despite the cause for concern, the new constitution offers some grounds for

optimism. As noted, several institutional changes in the 2010 constitution may help to

mitigate the divisive nature of political competition by devolving power to county

governments, strengthening the role of the legislature, and enhancing individual and

group rights. Needless to say, such ambitious constitutional provisions will only be

meaningful if implemented fully in coming years.

5. Conclusion

The central conclusion of this chapter is that power-sharing in the form of a grand

coalition can be a useful remedy for inter-communal violence in some contexts. In the

Kenyan case, few viable alternatives existed. While the opposition, believing that it had

rightly won the presidential election, preferred an immediate re-run of the election, it is

29

unlikely that PNU leaders would have accepted this option. Moreover, the escalation of

violence convinced key international actors – Kofi Annan and leading donor countries –

to prioritize ending the violence over concerns about the election outcome (Lindenmayer

and Kaye 2009; Brown 2009). Yet, while power-sharing may not have been anyone’s

preferred outcome ex ante, in the context of Kenya’s worsening crisis it provided an

attractive alternative. While important debates remain about whether power-sharing is a

useful institutional option in civil war settings, in Kenya power-sharing had an immediate

and positive effect, interrupting the escalating spiral of communal violence that followed

the 2007 election. It did so by coopting opposition leaders into the government and

signaling to opposition supporters that their communities would be represented.

Second, the Kenyan experience demonstrates that power-sharing – in the form of

a grand coalition – is unlikely to create incentives for political leaders to address the root

causes of conflict where doing so would threaten their own interests. In this regard,

expectations for power-sharing should be tempered: while it may serve as a useful

remedy in episodes of violence, it is unlikely to alter elite incentives and encourage

reform.

30

Acknowledgements An earlier draft of this paper was presented at American University in March 2009. I acknowledge several conference participants for valuable suggestions, especially Carl LeVan, Stephen Ndegwa, Karuti Kanyinga, Steffan Lindberg, Nicolas van de Walle, and Joel Barkan.

31

References Africa Confidential. “Kenya: Victory Is Not Enough.” 44(7): 3-5. April 4, 2003, Amadi, Henry. 2009. “Kenya’s Grand Coalition Government – Another Obstacle to

Urgent Constitutional Reform?” Africa Spectrum 44(3):149-164. Anderson, David and Emma Lochery. 2008. “Violence and Exodus in Kenya’s Rift

Valley, 2008: Predictable and Preventable?” Journal of East African Studies 2(2): 328-343.

Barkan, Joel D. 2008. “Will the Kenya Settlement Hold?” Current History, April 2008:

pp. 147-153. Barkan, Joel D. 1993. “Kenya: Lessons from a Flawed Election.” Journal of Democracy

4: pp. 85-99. Barkan, Joel D. and Makau Mutua. 2010. “Turning the Corner in Kenya.” Foreign Affairs, August 16. Barkan, Joel and Njuguna Ng’ethe. 1998. “Kenya Tries Again.” Journal of Democracy 9:

pp. 32-48. Bennett, George and Carl Rosberg. 1961. The Kenyatta Election: Kenya 1960-61.

London: Oxford University Press. Brown, Stephen. 2009. “Donor Responses to the 2008 Kenyan Crisis: Finally Getting It

Right?” Journal of Contemporary African Studies 27(3): 389-406. Brown, Stephen and Chandra Lekha and Sriram. 2012. “The Big Fish Won’t Fry

Themselves: Criminal Accountability for Post-Election Violence in Kenya.” African Affairs 111/243: 244-260.

Burgess, Robin, Remi Jedwab, Edward Miguel, Ameet Morjaria, and Gerard Padró i

Miguel. 2013. “The Value of Democracy: Evidence from Road Building in Kenya.” Working paper.

Cheeseman, Nic and Blessing-Miles Tendi. 2010. “Power-Sharing in Comparative

Perspective: the Dynamics of ‘unity government’ in Kenya and Zimbabwe.” Journal of Modern African Studies 48(2): 203-229.

Commission of Inquiry into Post Election Violence. 2008. Report of the Commission of

Inquiry into the Post-Election Violence. Nairobi.

32

Franck, Raphael and Ilia Rainer. 2012. “Does the Leader's Ethnicity Matter? Ethnic Favoritism, Education, and Health in Sub-Saharan Africa.” American Political Science Review 106(2): 294-325.

Gates, Scott, Kaare W. Strom, Benjamin A.T. Graham, Yonatan Lupu, and Havard

Strand. 2013. “Powersharing, Protection, and Peace.” Paper prepared for presentation at the Annual meeting of the American Political Science Association, August 29 – September 1, 2013, Chicago.

Gibson, Clark C. and James D. Long. 2009. “The Presidential and Parliamentary

Elections in Kenya, December 2007.” Electoral Studies 28(3): 497-502. Griffiths, Martin. N.d. “Prisoner of Peace: An Interview with Kofi Annan.” Center for

Humanitarian Dialogue. Hansen, Thomas Obel. 2013. “Kenya's Power-Sharing Arrangement and Its Implications

for Transitional Justice.” International Journal of Human Rights 17(2): 307-327. Harbeson, John W. 2012. “Land and the Quest for a Democratic State in Kenya: Bringing

Citizens Back in.” African Studies Review 55(1): 15-30. Horowitz, Jeremy. 2012. Campaigns and Ethnic Polarization in Kenya. Dissertation.

University of California, San Diego. Human Rights Watch. 1993. Divide and Rule: State Sponsored Ethnic Violence in Kenya.

New York: HRW. International Crisis Group. 2008. Kenya in Crisis. Africa Report No. 137. Jablonski, Ryan. “How Aid Target Votes: The Effect of Electoral Strategies on the

Distribution of Foreign Aid.” World Politics, forthcoming 2014. Kamugi, Prisca Mbura. 2009. “The Politics of Displacement in Multiparty Kenya.”

Journal of Contemporary African Studies 27(3): 345-364. Kanyinga, Karuti. 2007. “Governance Institutions and Inequality in Kenya.” In Society

for International Development, Readings on Inequality in Kenya: Sectoral Dynamics and Perspectives. Nairobi: SID Press.

Kanyinga, Karuti. 2009. “Hobbling along to Pay-offs: The Kenya Grand Coalition

Government.” Paper presented at the Conference on Political Inclusion, American University, Washington, DC, April 4-5, 2009.

Kanyinga, Karuti and James D. Long. 2012. “The Political Economy of Reforms in

Kenya: The Post-2007 Election Violence and a New Constitution” African Studies Review 55(1): 31-51.

33

Kanyinga, Karuti and Sophie Walker. 2013. “Building a Political Settlement: The

International Approach to Kenya’s 2008 Post-Election Crisis.” Stability: International Journal of Security and Development 2(2): 1-21.

Kenya Human Rights Commission. 1997. Kayas of Deprivation, Kayas of Blood:

Violence, Ethnicity and the State in Coastal Kenya. Nairobi: KHRC. Kenya Human Rights Commission. 1998. Killing the Vote: State Sponsored Violence and

Flawed Elections in Kenya. Nairobi: KHRC. Kenya National Commission on Human Rights. 2006. Referendum Report. Nairobi:

KNCHR. Kenya National Commission on Human Rights. 2008. On the Brink of the Precipice: A

Human Rights Account of Kenya’s Post-Election Violence. Nairobi: KNCHR. Kramon, Eric and Posner N. Posner. 2011. “Kenya’s New Constitution.” Journal of

Democracy 22(2): 89-103. Kramon, Eric and Daniel N. Posner. 2012. “Ethnic Favoritism in Primary Education in

Kenya.” Working paper. Kyle, Keith. 1999. The Politics of Independence Kenya. New York: St. Martin’s Press. Lijphart, Arend, 1977. Democracy in Plural Societies: A Comparative Exploration. New

Haven: Yale University Press. Lindenmayer, Elizabeth and Josie Lianna Kaye. 2009. A Choice for Peace? The Story of Forty-

One Days of Mediation in Kenya. International Peace Institute. Lynch, Gabrielle. 2006. “The Fruits of Perception: ‘Ethnic Politics’ and the Case of Kenya’s

Constitutional Referendum.” African Studies 65(2): pp. 233-270. Murray, Christina. Nd. Kenya’s 2010 Constitution. University of Capetown. Typescript. Ndegwa, Stephen. 2003. Kenya: Third Time Lucky?” Journal of Democracy 14(3): pp.

145-158. Njogu, Kimani, ed. 2009. Healing the Wound: Personal Narratives about the 2007 Post-

Election Violence in Kenya. Nairobi, Kenya: Twaweza Communications Ltd. Sanger, Clyde and John Nottingham. 1964. “The Kenya General Election of 1963.” Journal of

Modern African Studies 2(1): pp. 1-40.

34

Timothy D. Sisk. 1996. Power Sharing and International Mediation in Ethnic Conflicts. Washington, D.C.: United States Institute of Peace.

Throup, David. 1987. “The Construction and Destruction of the Kenyatta State.” In Michael

Schatzberg, ed., The Political Economy of Kenya. New York: Praeger. Throup, David. 2008. “The Count.” Journal of Eastern African Studies 2(2): pp. 290-304. Throup, David and Charles Hornsby. 1998. Multi-party Politics in Kenya. Oxford: James

Currey. Walter, Barbara. 1997. “The Critical Barrier to Civil War Settlement.” International

Organization 51(3): 335-364. Newspaper Articles British Broadcasting Corporation. “Kenya Rivals Agree to Share Power.” February 28,

2008. http://news.bbc.co.uk/1/hi/world/africa/7268903.stm (October 9, 2013). Business Daily. “Commission Lacks Power to Resolve Land Issues, Says Ndung’u.”

March 24, 2013. http://www.businessdailyafrica.com/Commission-lacks-powers-to-resolve-land-issues/-/539546/1729304/-/12bseixz/-/index.html (October 8, 2013).

Kabaria, Judie. 2013. “Kenya: Too Little Action on Hate Speech?” Institute for War and

Peace Reporting, July 11. http://iwpr.net/report-news/kenya-too-little-action-hate-speech (October 8, 2013).

Namwaya, Otsieno. 2004. “Who Owns Kenya.” East African Standard, October 1.

http://blog.marsgroupkenya.org/2008/02/04/who-owns-kenya/ (October 8, 2013). Malalo, Humprey. 2013. “Kenya’s Ruto Loses Land Grab Case.” IOL News, June 28.

http://www.iol.co.za/news/africa/kenya-s-ruto-loses-land-grab-case1.1539378#.UkVhWLwwiQI (October 8, 2013).

35

Table 1. Provisions of the Power-Sharing Agreement Unanimity rules None in legislative arena. In the Cabinet,

appointments and removals must be approved by both parties.

Grand coalition Yes Fixed equality of representation among different groups

No, except that Cabinet positions and top administration posts will be shared evenly by the two main political parties. No specific provisions for particular ethnic or religious groups.

Proportionality in election results No. Kenya retains a presidential system with first-past-the post constituencies for Parliament.

Proportionality in gov’t appointments, contracts, rewards

Yes, only with regard to higher-level appointments. No specific provision for lower-level appointments in civil service, contracts, or other rewards.

Federalism No Group autonomy provision, for example with respect to legal regulations, education, or religious matters

No

Guarantees of individual liberties No new provisions added above those already contained in the existing constitutions.

Power-dividing rules (restricting the gov’t from regulation of certain policy areas, such as religious practices, language)

No

Wealth-sharing No Sunset (temporary) provisions with respect to any aspects of power-sharing

Yes. The coalition government will be dissolved if the tenth parliament is dissolved; if the parties agree in writing; or if one coalition partner withdraws from the coalition.